| dc.description.abstract | High fertility in any part of the world is a cause for concern among policy makers. In subSaharan Africa, high fertility has been attributed to unwanted childbearing and a desired family size of above two surviving children. Kenya's fertility history can be traced from the 1960s when family planning programmes were initiated. During this period, the total fertility rate (TFR) had been increasing steadily and was the highest in the world at 8.06 in the mid 1970s. This trend was however reversed from the 1980s till 1998, when TFR declined from 6.7 to 4.7. In 2003, fertility stalled at a high level of 4.9 children per woman. There is a need to explain this stall so as to attain national goals and targets on fertility.
The study of the determinants of birth intervals accounts for changes in the timing of childbearing and thus gives a better picture than the study of the determinants of completed fertility. The general objective of this study was to identify the determinants of birth intervals in Kenya. Specifically, it sought to establish the determinants of birth intervals among married women in Kenya as well as examining the role of intermediate, demographic, socio-cultural and economic factors that influence birth intervals in the country.
Data for this study was drawn from the 2003 KDHS where a sample of 2608 women, who reported being currently married and also in first union, were selected. A total of 5216 birth intervals were obtained, this being a combination of 2608 open and 2608 closed intervals. Frequencies provided a description of the study population while the cox proportional hazards model was used for the bivariate and multivariate analysis.
At the bivariate level, only sex of the index child was found to be insignificant relative to other independent variables considered in the study. All other variables had a statistically significant association with the risk of giving birth. At the multivariate level, three models were fitted to assess the effect of the independent variables on the risk of giving birth. Model one comprised of demographic variables; model two introduced socio-economic and cultural variables while model three added intermediate variables.
In all three models, sex of the index child was statistically insignificant. This was the only insignificant variable in model one. In model two, all variables except maternal education and sex of the index child were statistically significant. However, paternal education was only statistically significant for those who had attained secondary and above levels of education. In model three, the effect of the background factors were diminished with the introduction of the intermediate variables. As such, only age of the mother at the start of the interval, birth order, ever use of contraception and duration of breastfeeding had statistical significance.
The study concluded that the following are important covariates of the risk of giving birth; age of the mother at the start of the interval, birth order, survival status of index child, length of preceding birth interval, maternal and paternal education, region, type of place of residence, ever use of contraception and duration of breastfeeding. The study recommended the promotion of contraceptive use and breastfeeding practices due to their association with a reduced risk of giving birth. There is also a need to promote universal education as a step towards tackling high fertility. The study also recommended further research on the topic to shed light on the link between birth intervals and completed family size. | en_US |
